Every Child Wins A Prize: Federal Seats With Swings To All Contestants

During last night's Cook by-election count there were a few comments about the swing column.  All six parties/independents had recorded a positive swing from the 2022 election.  In the case of Cook this was not at all surprising - three of the parties and the one independent had not even run in the seat in 2022, so their "swing" was automatically plus.  The Greens were always going to get a primary vote swing with no Labor candidate and no prominent left/centre independent.  That left the Liberals, and the question was whether they could gain enough primaries from the 34.6% who voted Labor, UAP or One Nation in 2022 to compensate for replacing a former Prime Minister and 17-year incumbent with some dude from outside the electorate.  This they did with 7% to spare and lo and behold there's a neat little line of pluses in the swing column for the recontesting candidates:

(Apologies to AEC, I've pinched the Wikipedia version for clearer display)

This is a common event in by-elections where one major party doesn't contest.  It has happened by my count in 9 of 21 such by-elections in the last 50 years, the others being Perth and Batman 2018, Higgins 2009, Isaacs 2000, Holt 1999, Blaxland 1996, Wentworth 1995, and Menzies 1991.  Perth 2018 achieved this feat despite having 15 candidates, however only three parties were recontesting.  Blaxland 1996 had five recontestants - I should note that I treat an independent as such only if it is the same person running and doing so as an independent both times.  

The event sometimes happens at by-elections where both major parties do contest, but very rarely these days - I get twice in 43 chances in 50 years (it was somewhat more common before that).  Groom 2020 was the most recent - with One Nation, Greens, UAP and FACN all not recontesting after polling 28% between them, the majors picked up primary vote swings of 6.5% (LNP) and 8.6% (Labor).  The only minor parties contesting Groom were Liberal Democrats and Sustainable Australia.  The other case in the last 50 years I found was Moreton 1983.  The Australian Democrats had polled 6.7% at the general election and did not contest the by-election; the two minor candidates who did polled less than 1% apiece, and both major parties got a primary vote swing.

But how often, I thought, does this happen in a seat contest at a general election?  The most obvious cases would be cases where a high-profile independent who ran one election did not run in the next, so my first two guesses were New England 2013 (Tony Windsor vacating) and Lyne 2013 (Rob Oakeshott vacating), and these were correct.  But I didn't think it would be common, since both major parties contest every seat and usually there will be a swing against at least one.  It turned out there were some recent elections where the event was more common than I thought.  

The feat becomes more likely the better the federal primary swing results for the major parties,  because if one of the major parties has a big primary vote swing against it, it is likely that there will be a primary vote swing against it in nearly every seat.  It also becomes more likely when a minor party that polled substantially at the previous election either doesn't run at all or runs in fewer seats.  However a minor party recontesting many seats but with a primary vote swing away from it will stop the feat in many divisions.  Also, as well as vacating incumbents there's another case where this event becomes more likely to occur, and that's where the Coalition parties ran in a three-cornered contest at the previous election.

2022 (1 seat)

There was a large primary vote swing against the Coalition, and as a result only five seats had swings to both/all major parties contesting (the "all" is for three-cornered contests).  Of these Hunter had a huge swing against One Nation, and Fraser (UAP), Dawson (KAP, UAP) and O'Connor (ON, UAP) were wrecked by small swings against minor right parties.  The sole survivor was Mallee which very barely recorded a swing to all eight contestants (five recontesting), coming off a three-cornered contest with 16 candidates in 2019.  The swing to Labor in Mallee was 0.22%.

2019 (10 seats)

There were seven seats in South Australia with swings to everybody, all coming off the disappearance of Nick Xenophon Team/Centre Alliance from their ballots: Adelaide, Boothby, Hindmarsh, Kingston, Makin, Spence and Sturt.   Three others ticked the box: New England (no Tony Windsor who ran but lost in 2016), Holt and Menzies.  In Holt the UAP polled less than the combined might of Family First and Rise Up Australia had in 2016, and both majors and the Greens had a swing.  In Menzies five units good for 14.6% between them in 2016 (most prominently independent Stephen Mayne) had not stumped up again and the DLP, UAP and Reason were only good for 7.8%.

2016 (18 seats)

This is the most productive recent election by far for this feat with 18 cases, mainly because the Palmer United Party which polled 5.5% in 2013 disappeared from every seat but one.  Also the primary vote swing against the Coalition (3.5%) was not that large, and the National Party did quite well.  

NSW: Hunter, Lyne, Parkes, Robertson, Riverina, Werriwa

Vic: Mallee, Casey, Gorton, Isaacs, Lalor

Qld: Bowman, Capricornia, Fairfax, Fisher, McPherson

WA: Moore

ACT: Canberra

Here Mallee was coming off a three-cornered contest and Lyne and Robertson had had significant independents in 2013 (giving Lyne two in a row).  In the case of Lalor there had been 11 candidates in 2013 (not sure whether this had anything to do with PM Gillard or her departure from the seat) and in 2016 there were only five.  

2013 (2 seats)

The cases of New England and Lyne were mentioned above.  But they are the only two, because the fairly large primary vote swing against Labor left just five other seats with a swing to both major parties. Those were the inner-city set of Grayndler, Sydney, Ryan and Perth, and also Chifley where Ed Husic must have done something right.  But 2013 also saw a national swing against the Greens, including in all those five seats.

2010 (0 seats)

Nothing!  2010 is the only election since the introduction of the current party registration system at which I find no seat with a swing to all contestants.  A 5.4% primary vote swing against Labor meant very few seats recorded swings to Labor at all, and of these only one (Corio) recorded a primary vote swing to the Coalition.  Corio was coming off a 12.7% vote for Gavan O'Connor, who was dumped by Labor in 2007 and ran as an independent.  However, its bid for solitary positive swing glory was foiled by a 0.3% swing against Family First.

2007 (2 seats)

Just two, Calare (Peter Andren vacancy) and McMillan (coming off a three-cornered contest).  There were another ten seats with swings to both major parties - many of those without any obvious candidate factor driving it - but all those were ruined by swings against various minor parties, including in most cases the Citizens Electoral Council.  There as also a very small swing against Bob Katter in Kennedy.

2004 (5 seats)

2004 has the issue that while neither major party did badly in primary vote swing terms, both the Democrats and One Nation were on the skids but still running plenty of candidates, therefore in most cases one or both of them had a swing against even if the major parties didn't.  The most prospective seats therefore were seats that neither contested, and these five duly recorded swings to everyone: Page, Ballarat, Bendigo, Corangamite and Franklin (the latter is the only Tasmanian case I've found since 1983).  Page was coming off a very large field including a significant independent.  

2001 (1 seat)

Another year that looks good because of benign major party swings, but the swing against One Nation off their 1998 high affects a lot of seats.  The sole winner was Menzies, which was not contested by One Nation and was coming off a significant independent run by euthanasia campaigner Philip Nitschke against Kevin Andrews.

1998 (1 seat)

A huge primary vote swing against the Coalition (7.7%) made it hard for every candidate to get a swing anywhere in 1998, the only example being Wills (no Phil Cleary).  There were another ten seats with swings to both major parties but Curtin and Moore had the defeat of sitting independents, another seven had swings against the Democrats and in Melbourne there were swings against the Greens, Natural Law and, alas, the Imperial British Party.  

1996 (2 seats)

Another election with a big primary vote swing against one side, in this case 6.2% against Labor.  Everyone got the goodies in North Sydney (on the retirement of Ted Mack) and Throsby (in the absence of Rex Connor who had polled 10.5% in 1993).  The only other seat with a swing to both majors was Denison, which had a swing against the Greens after Bob Brown ran for the seat in 1993.

1993 (5 or 6 seats)

1993 seems promising because there were swings to both major parties after the 1990 election which had been a low water-mark for the major party vote.  However these swings came mainly off the Democrats, and so swings to everyone were only likely in seats that the Democrats did not contest.  There were five of them in NSW: Berowra, Gwydir, Lindsay, Macarthur and North Sydney.  However North Sydney doesn't really count as the swing to independent Ted Mack is from a redistributed baseline that assigned him zero votes in areas that had been outside the seat in 1990.  In raw terms there was a swing against him and there would have been within the areas common to both elections.  A sixth case was Northern Territory (for the second election in a row) as there were only two candidates compared to six.  

1990 (2 seats)

The 1990 election saw swings against both major parties, including a large swing against Labor.  The only cases of swings to everyone were Prospect (coming off an independent Sam Barone getting 14.8% in 1987) and Northern Territory (coming off a three-cornered contest).  

1987 (5 seats)

1987 seems prospective as there was a small swing against Labor and a small swing to both Coalition parties.  However there were a lot more three-cornered contests in those days (this is the Joh for Canberra election) and it's harder to get swings to all of Liberals, Nationals and Labor in the same seat.  The only seats with swings to everyone were Greenway, Mitchell, Riverina-Darling, Wannon and Cowan.  Riverina-Darling and Wannon were both coming off three-cornered contests while Cowan was in fact, unusually, going onto one (if one can call it a contest - the National in Cowan polled less than the 1984 Democrat.)  The first three listed had only two candidates each.

1984 (2 seats)

1984 also seems prospective for the same reason as 1987, but there are similar issues.  Another problem with 1984 is that for both/all major parties to get a swing, the swing needs to come from somewhere, and outside the majors and Democrats (whose result was pretty even) there was only 1.9% of vote nationwide to draw from.  The only seats with all positive swings were Barton and Lyne.

Before 1984

The further back one goes the messier it gets in terms of identifying parties, "Independent Liberal", uncontested seats and so on; these are some draft lists from skimming the Wikipedia pages that are given just to give an idea of how common a swing to everybody might have been in different elections.  

1983 (5): Lyne, Riverina, Corangamite, Griffith, Maranoa

1980: None

1977 (4): Cowper, Farrer, Ryan, Hindmarsh

1975: None (Swing against Labor in every seat)

1974 (8): Banks, Kingsford-Smith, Prospect, Reid, Shortland, Werriwa, Herbert, Denison

1972 (3): Banks, Mallee, Wimmera

1969 (3): Hunter, Lalor, Oxley

1966 (1): Hunter

That's as far back as I feel inclined to go for now.  Corrections welcome, there may well be errors somewhere in this piece.  

2025?

An interesting thing for 2025 will be whether the United Australia Party runs again (presumably under a different name) or not.  If they don't, and if primary vote swings involving the majors are relatively small, then 2025 could be another election with a number of cases of this feat.  If UAP do run again, then large fields in many seats could make it hard for cases of everyone getting a plus swing to occur.  I will update this article sometime after that election.  

One can of course also do this for state elections.  In Victoria 2022 it happened in Geelong, Pascoe Vale and arguably Ovens Valley (all coming off independent runs - in Ovens Valley Labor's swing was zero to one decimal but probably positive before rounding).  In NSW 2023 it happened in East Hills and Wollongong (each had three fewer minor candidates than 2019) - there were two other near misses with tiny swings against the Greens.  In Queensland 2020 it happened in Callide (no One Nation candidate; large swings against One Nation stopped it in most other seats.) SA 2022 was coming off a SA-Best run and swings to everyone occurred in 20 out of 47 seats!  In WA 2021 it didn't happen at all, nor in WA 2017 or WA 2013; there were two cases in WA 2008 (Girrawheen and Willagee).  In NT 2020 it happened in Arafura, Karama, Mulka (no CLP candidate) and Wanguri.  Tasmania has only five multi-member divisions and the last cases of the feat at party level were Bass and Franklin 1986 (both coming off independent runs).  In the ACT, which also has multi-member divisions, it has never happened.